# CMap issue in PDF file generated with recent TeX Live + Ghostscript

Marcel Fabian Krüger tex at 2krueger.de
Sat Oct 30 01:29:12 CEST 2021

```On Fri, Oct 29, 2021 at 12:07:03PM +0200, Vincent Lefevre wrote:
> On 2021-10-19 19:22:04 +0200, Ulrike Fischer wrote:
> > LaTeX doesn't only activate \pdfgentounicode=1, it also load
> >
> > \pdfglyphtounicode{Aogonek}{0104}
> >
> > then after ps2pdf the cmap for fi is wrong again.
>
> Is there a way to avoid the load of glyphtounicode (while still
> allowing \pdfgentounicode=1 for users who need to add their own
> \pdfglyphtounicode entries) so that the obtained PDF conforms to
> the original definition of ToUnicode CMaps?
>

In order to get a good Unicode mappings, mapping ligatures and similar
glyphs which are most likely to cause issues in the PDF reader is the
most important aspect. So disabling the mapping for these would be the
wrong thing. If you use a tool which does not support standard
mappings for them, you have to decide which codepoint to map them to